assemble_predictions: Assemble leave-one-season-out or test phase predictions made...

In reichlab/adaptively-weighted-ensemble: Adaptively Weighted Ensembles via Stacking

Description

Assemble leave-one-season-out or test phase predictions made by kde, kcde, and sarima models on training data.

Usage

assemble_predictions(preds_path = "inst/estimation/loso-predictions",
  regions = c("National", paste0("Region", 1:10)), seasons = "*",
  models = c("kde", "kcde", "sarima"), prediction_targets = c("onset",
  "peak_week", "peak_inc", "ph_1_inc", "ph_2_inc", "ph_3_inc", "ph_4_inc"),
  prediction_types = c("log_score", "competition_log_score", "bin_log_probs"))

Arguments

preds_path	path to leave-one-season-out or test phase predictions.
regions	character vector specifying regions for which to get predictions, "National" or "Regionk" for k in 1, ..., 10.
seasons	character vector specifying seasons for which to get predictions, "2011/2012" or "2011-2012" or "*" for all seasons
models	character vector specifying models for which to get predictions, "kde", "kcde", or "sarima"
prediction_targets	character vector specifying prediction targets, "onset", "peak_week", "peak_inc", "ph_1_inc", ..., "ph_4_inc"
prediction_types	character vector specifying prediction types, "log_score", "competition_log_score", or "bin_log_probs"

Value

a data frame with predictions.

reichlab/adaptively-weighted-ensemble documentation built on May 27, 2019, 4:53 a.m.